We aim to elucidate the molecular mechanisms that control the assembly of viral capsids with the ultimate goals of defining targets for antiviral compounds and of gaining insight into the mechanisms that control the assembly of macromolecular complexes in general. In particular, we focus on the large-scale conformational changes that accompany capsid maturation and interaction of virions with host cells. Our major progress over the past year was as follows: (i) Hepatitis B Virus Capsid ("Core Antigen"). Despite the availability of effective vaccines, HBV remains a public health problem of immense proportions, motivating further efforts to elucidate its replicative cycle. We have been studying its capsid protein (183 amino acids), which we first found to form dimers that are capable of self-assembly into icosahedral particles. These particles are of two different sizes - one with T=3 symmetry (90 dimers; 28nm diameter), the other with T=4 (120 dimers; 34nm diameter). Unlike most other viral capsid proteins, that of HBV was found to be predominantly alfa-helical. In 1997, we calculated a density map from cryo-electron micrographs at the unprecedentedly high resolution of 0.9nm, in which much of the secondary structure was directly visible, including the 4-helix bundle that forms the dimerization motif. Our continuing research has aimed at establishing the precise locations of particular amino acid residues to help delineate the overall path of the polypeptide chain through our density map. Our first success, reported in last year's report and since published, was to localize the C-terminus by gold-cluster labelling. More recently, we have pinpointed a surface loop comprising residues 78-83 by characterizing the binding of Fab fragments of a monoclonal antibody that recognizes this peptide. Localization of the N- terminus was also accomplished by appending an extraneous octapeptide at this site, which we then visualized as additional density in a difference map. We are continuing to extend the resolution of our map of the capsid and to impose more constraints for chain-tracing. (ii) Structure and Tegumentation of the Cytomegalovirus Capsid. Cytomegalovirus (CMV) is a clinically important member of the herpesvirus family, causing infections in immunosuppressed individuals. It is also unresponsive to the antiviral, acyclovir, which is effective against other herpesviruses. In this context, we have been studying cytomegalovirus capsid assembly, using simian CMV as a model for the similar but less tractable human CMV. The main thrust of this project over the past year has been to exploit the opportunity offered by the SCMV system to study capsid-tegument interactions. The tegument is an extensive compartment of proteins situated between the herpesvirus capsid and its envelope. Despite longstanding interest, the molecular architecture and functionale rationale of the tegument have remained obscure. At late times after the infection of cultured fibroblasts with SCMV, tegumented capsids appear in the cytoplasm. By isolating these capsids and comparing their structure and protein composition with those of conventional nuclear capsids, we have characterized the modes of binding of two tegument proteins and identified provisional candidates for them as the "basic phosphoprotein" (119 kDa) and the "upper matrix protein" (69 kDa), respectively. These observations represent the first identification of such linkages for any herpesvirus. (iii) Maturation Dynamics of Bacteriophage HK97. Our primary interest in bacteriophage capsid assembly lies in the monumental conformational changes that accompany maturation of precursor capsids. These changes are irreversible, frequently involve partial refolding of the subunits, and are stringently controlled. As such, they afford unique opportunities for insight into large-scale conformational changes. HK97 represents an advantageous system to study these reactions. The earliest precursor, Prohead I, is converted to Prohead II by proteolysis, then expands to Head I, facilitating covalent cross-linking of Head II, the end-state. Our first study on HK97 compared the 3- dimensional structures of all four particles at a resolution of ~ 2.5 nm. We have since gone on to investigate the dynamic progression of expansion (Prohead II -< Head I) by inducing this event in vitro at low pH, and monitoring subsequent events by time-lapse cryo-electron microscopy. Analysis of the immense volume of data generated in these experiments is proving to be time-consuming but rewarding. Last year, we reported our first results whereby a semi-expanded particle was observed after ~ 30 minutes. We have now characterized intermediates from three later time points and succeeded in obtaining a reconstruction of the spherical "balloon" particle that predominates after 3 hours. Our current account of the acid-induced maturation pathway is that a semi-expanded state is reached rapidly, i.e. within a few minutes; thereafter there is little change in size for ~ 2 hrs, although small changes in capsomer shape continue to take place; then, these intermediates switch into the "balloon" particle; finally, on restoring to neutral pH, the polyhedral Head I state is achieved. A movie has been made of this dynamic process, using the graphical technique of "morphing" to connect experimentally visualized states. (iv) Conformational Changes in Poliovirus upon Interaction with Host Cell. When picornaviruses such as the human pathogen, poliovirus type 1, encounter susceptible cells, the viral capsid binds to a cellular receptor. Ensuing events leading to the intracellular propagation of the infecting virus are not well understood. However, with poliovirus, interaction with its receptor induces a conformational change in the virion, as manifested by a change in its sedimentation coefficient from 160S to 135S. RNA is then released, leaving an 80S capsid. We are investigating the structural basis of these transitions, which may be simulated in vitro by brief heating in an appropriate buffer. The 160S virion has been solved to high resolution by X- ray crystallography, but the 135S and 80S states have not been conducive to such analysis.